Next Generation Microservicesbased Data and Analytics Solution

Transcription

1 Next Generation Microservicesbased Data and Analytics Solution Gaurav Kakad Solution Architect, Tata Consultancy Services Bryan Werwick Solution Architect, Tata Consultancy Services May 9, 2018

2 Agenda Business Problem Our Solution Benefits & Features Technology Landscape & Architecture Demo

3 Impact of Big Data & Analytics The biggest impact areas of big data was in customer relationship management 40% companies worldwide analyze big data 8% Increase in revenues Research by Forrester 10% Survey by BARC A large American bank is using sophisticated predictive models to analyze historical transactions and 115 variables to forecast customer churn. The company believes it can now accurately identify 24% of accounts within its Australian branch that will close within the next four months and can take required steps to prevent them from churning. 50% Applications used by Fortune 500 organizations are already on the cloud Survey by IDG Reduction in Costs Survey by KPMG There will be a rise in demand for analytics tools that are simple, flexible, and capable of handling a variety of data sources. 10% Healthcare professionals use advanced data analytics tools Analyzing big data helped a large printer manufacturer to cut the attrition rate at their call centers by over 20% a significant and tangible financial saving. A large local bank by market capitalization in Singapore that operates in 15 countries world wide was able to achieve higher customer engagement and increase customer satisfaction by 20% compared to a control group. The bank was able to benefit by responding to the customer actions, personal lifetime events and demographic profiles.

4 Business Problem Different marketing units working in silos and unavailability of social media profile of a customer created a half-matured marketing engine, unsatisfactory customer experience and loss of business opportunities. Roadblocks for business were: Unfocussed marketing efforts Disparate Business Intelligence units Unconsolidated data & inaccurate insights Extended cycle time to process data Lack of coordination between big data teams Dependency on legacy systems to process data Non-standard governance process for data Reasons for unsatisfactory customer experience were: Unsolicited promotions/offers Spam s/multiple communication Brand loyalty benefits not streamlined Slow complaint resolution Lack of 360 degree view of customer Lack of personalized experiences or offers Data silos, multiple data sources for analytics Product & channel centric business processes Current technology unsuitable for scaling and processing social media data

5 Our Solution A Microservices-based Big Data and Analytics solution that will help you to understand your customer better. It is based on the Hybrid Cloud and Container Framework. It will provide a unified view about your customers by blending their transactional and digital profile. You can gain deep insights about your customer, through a single platform. Blends customer s digital and transactional persona Shows real-time dashboard view Tracks customer s digital foot prints and identifies hotspots for cross/up sell Forecasts future trends, sends alerts, notifications, and initiates call for actions Captures customer information through Clickstream, survey, content and social media analytics Determines Next Best Offers (NBO), Next Best Actions (NBA) and Next Product to Buy (NPTB) models Product Account Channels Persona Customer 360 Lifetime Events Sentiments Household Influencers Product vs Channel Affinity

6 Benefits Innovative Statistical Models Creating better customer experience Improved lead generation & customer service Reduced marketing & service cost Increased ROI by data driven offerings & targeted marketing Better and informed decisions Sentiment Analysis Multi -Channel Analytics Customer Intelligence Campaign Metrics Single View of Customers Predictive Analysis

7 Benefits Powered by Red Hat OpenShift Container Platform, it provides consistent application development and deployment platform for hybrid cloud. Enterprises can leverage Agile and DevOps methodologies and technologies leading to accelerated time-to-market for new applications and services. It is a Microservices-based application architecture. Enables dynamic scaling requirement of the enterprises. Improved Agility & Scalability through pay-as-you-go cloud based model. Increased profitability by leveraging data-driven insights to expedite decisionmaking, increase focus on customer-centric business processes, and maximize customer lifetime value. The solution harnesses an Open Source centric and continuous innovation led architecture around Cloud and Machine Learning and is relevant for all industries across both horizontal as well as industry micro-vertical specific use cases.

8 Features Machine Learning, Advanced Predictive Analytics & Recommendation System Allow organizations to analyze and correlate data generated at every stop of the customer s journey. The collected information can be accessed and analyzed at any given time, to gain systematic understanding of the customer s sentiments. It will help to streamline operations by predicting consumer behavior and reduce churn through Advanced Predictive Analytics and Recommendation Systems. Advanced Dashboard A single dashboard will provide all business KPI s. It will be able to send alerts, notifications and initiate call for action. Ready framework for Data Ingestion & Management All the data from core applications and social media channels will be collected automatically. Data will be governed by automated and mandatory meta - data management system. Security & Controlled Access in real-time The solution will offer role based user access to the applications. It will provide secured remote login and encrypted communication between the client and the applications. Easily integrated system, on the cloud Now you will be able to schedule jobs, provision and manage data with an easily integrated system. It will allow you to make choices about Private vs Public Cloud in terms of Data Residency and still have consistent developer and business user experience.

9 Business Drivers

10 Technology Landscape Red Hat Suite of Products OpenShift Container Platform JBoss Data Virtualization JBoss AMQ Cloud Mgmt. & Automation Red Hat Operations Network Ansible Tower 3-Scale AMP GlusterFS Spark Red Hat JDV OpenShift MongoDB Other Tools / Software PostgreSQL MongoDB Python Apache Spark Python JBoss AMQ Gluster PostgreSQL

11 Batch Batch Mode Real Time Streaming Mode Architecture Data Ingestion Layer Data Processing /Storage Layer Clickstream Social Media Red Hat JBoss A-MQ Real-time Processing External BI Tools Files Surveys s Red Hat JBoss DV S3 API Batch Processing Red Hat JBoss Data Grid Red Hat JDV 3 Scale JDV BI Interface Applications RDBMS Scheduler

12 Use Case

13 Flashback to before Customer Analytics Can I get information across all the Customer touch points? Is it possible to get better insight from introducing Customer feedback along with Operational data? Predictive & Prescriptive Pre-deployment scenario Communication Social Will it be possible to provide futuristic view for some key business indicators? Clickstream Survey

14 The Problem Statement Aaron is a long-term customer of the bank, with multiple accounts. He likes the convenience of online services but becomes frustrated when things don't work as easily or as quickly as he'd like. In such cases, he browses through other Bank s site to get better option. He doesn t hesitate to drop a mail to the customer service or write comment on Bank s social handler/ page when he has a problem and has done so on a couple of occasions in the last month. Aaron

15 Portfolio Manager Start of the Day Can I get information across all the Customer touch points? Is it possible to get better insight from introducing Customer feedback along with Operational data? Will it be possible to provide futuristic view for some key business indicators?

16 Portfolio Manager After 2 attempts to complete his bank's online Personal Loan application, Aaron gives up and drop mails to Customer service and write comment on Bank s social handler/ page. Through his segmentation analysis, the Portfolio Manager finds out Aaron's demographic, transactional, social, clickstreams and other behavioral data points and able to see his failed attempts to submit his Personal loan application; he also sees that: Aaron recently dropped mail to Customer service as well as lodge complaint through social media about difficulty with other online services Browsed on web platform on Personal loan of other Banks' and compared the rates

17 Portfolio Manager The combined issues alerts the Portfolio Manager that Aaron's satisfaction score has declined as well as Churn Propensity has increased which gives an alert that Aaron may be a high Churn risk. At the same time, Portfolio Manager also sees that Aaron is a high-value Customer and is prompted onscreen to escalate Aaron s complaint to accelerate issue resolution.

18 Next Best Offering i. Connect Aaron via video chat to one of Bank's Personal Loan officers to complete his Loan application without any further difficulty ii. Uncovers new customer insight by analyzing web behaviors, which helped guide development of more personalized offers and marketing outreach

19 Demo - How we achieved this?

20 Customer Transactions Landing Zone (CSV Files) GlusterFS Customer Summary

21 Sentiment Analytics Twitter User Details Twitter API Sentiment Calculation Read Properties From GlusterFS GlusterFS Write Twitter Sentiments and Details in GlusterFS

22 Sentiment Analytics API Sentiment Calculation Write Text data Into GlusterFS Read Properties and Text data From Gluster GlusterFS Write Sentiments and Details in Gluster

23 Sentiment Analytics Customer Surveys PDF to Text Sentiment Calculation Landing Zone (PDF Files) Read PDF file from GlusterFS Write Text to GlusterFS GlusterFS Read Properties and PDF Text data From GlusterFS Write Sentiments and Details in GlusterFS

24 Data Integration Customer Profile Customer Transactions Sentiment Data Integration Red Hat JBoss Data Virtualization OData Survey Sentiment Twitter Sentiment

25 Q&A

26 THANK YOU plus.google.com/+redhat facebook.com/redhatinc linkedin.com/company/red-hat twitter.com/redhat youtube.com/user/redhatvideos